Due to the relative paucity of nephrologists, primary care providers (PCPs) deliver most of the care to patients with chronic kidney disease (CKD). However, PCPs do not recognize the presence of CKD as well as nephrologists nor are they aware of existing kidney disease guidelines. These factors contribute to inadequate outcomes in CKD patients. Yet, very few studies have examined methods to improve PCP care of CKD patients. Computerized clinical decision support systems (CDSS) are an effective tool for enhancing PCP provision of care in other settings, but their role in CKD care has not been established. The long-term objectives of the application are: 1)To characterize current PCP practice patterns in the care of CKD patients and 2)To improve CKD outcomes by increasing PCP recognition of CKD and optimizing PCP compliance with evidence-based CKD guidelines through the use of CDSS. The specific aims are 1). To determine the extent that PCPs recognize CKD 3b-5 (eGFR<45ml/min/1.73m2) before and after the institution of routine eGFR reporting. 2). To ascertain the proportion of patients with CKD 3b-5 who (a) had a urinary spot albumin to creatinine ratio (ACR) in the past year, (b) are on an angiotensin converting enzyme inhibitor (ACE) or angiotensin receptor blocker (ARB) prior to and following routine eGFR reporting. 3). To demonstrate the effect of a randomized controlled trial of PCP education and clinical reminders compared to provider education alone to increase the proportion of patients with CKD 3b-5 referred to a nephrologist. The proposal consists of a two-phase clinical study among outpatients with CKD 3b-5 at a single academic clinical site. In the first phase, data will be abstracted from electronic medical records (EMR) to determine the proportion of CKD 3b-5 patients with a diagnosis code for CKD, ACR, and ACE/ARB treatment before and after the institution of routine eGFR reporting. In the second phase, a 6 month pilot randomized controlled trial of PCP education and EMR clinical reminders versus education alone will compare the proportion of CKD 3b-5 patients with a nephrology referal, ACR, and ACE/ARB use. The intervention consists of education and automated EMR reminders to refer to nephrology and measure ACR in CKD 3b-5. Kidney disease is an increasingly common disorder but due to the scarcity of kidney doctors, primary care physicians treat many kidney disease patients without the help of a specialist. Studies have shown that the care delivered to these patients is often suboptimal. This proposal will investigate whether computerized alerts that automatically remind doctors when their patients have significant kidney disease and suggest interventions that should be ordered will improve the care delivered to patients with chronic kidney disease.